Add reward-modulated learning and adaptive sleep budget controls#4
Merged
Add reward-modulated learning and adaptive sleep budget controls#4
Conversation
Introduces optional reward-based scaling of wake learning rates and adaptive sleep budget scaling to circadian predictive coding models for both NumPy baseline and ResNet variants. Addresses model review feedback by enabling task-relevant learning prioritization and reducing manual tuning of sleep structural budgets. Adds CLI/config flags for control, updates tests for coverage, and documents new mechanisms. Improves model flexibility while keeping core behavior deterministic and opt-in.
Improves out-of-the-box model behavior by setting adaptive sleep budget scaling as the default for ResNet benchmark runs. Updates CLI options for clearer enable/disable semantics and aligns documentation to reflect new defaults. Promotes more robust performance in circadian predictive coding experiments.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary\n- add reward-modulated wake learning and adaptive sleep budget scaling to NumPy circadian core\n- extend the same controls to the ResNet circadian head and benchmark CLI/config\n- add tests for reward scaling and adaptive budget behavior (NumPy + Torch paths)\n- set ResNet benchmark default to adaptive sleep budget enabled, reward modulation disabled\n- update docs and ADR/review notes\n\n## Validation\n- ruff check .\n- mypy src tests scripts\n- pytest -q\n- additional CIFAR-100 A/B and active-sleep comparisons run during development